Electrochemical ohmic memristors for continual learning

Abstract Developing versatile and reliable memristive devices is crucial for advancing future memory and computing architectures. The years of intensive research have still not reached and demonstrated their full horizon of capabilities, and new concepts are essential for successfully using the comp...

Full description

Saved in:
Bibliographic Details
Main Authors: Shaochuan Chen, Zhen Yang, Heinrich Hartmann, Astrid Besmehn, Yuchao Yang, Ilia Valov
Format: Article
Language:English
Published: Nature Portfolio 2025-03-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-57543-w
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850251517547773952
author Shaochuan Chen
Zhen Yang
Heinrich Hartmann
Astrid Besmehn
Yuchao Yang
Ilia Valov
author_facet Shaochuan Chen
Zhen Yang
Heinrich Hartmann
Astrid Besmehn
Yuchao Yang
Ilia Valov
author_sort Shaochuan Chen
collection DOAJ
description Abstract Developing versatile and reliable memristive devices is crucial for advancing future memory and computing architectures. The years of intensive research have still not reached and demonstrated their full horizon of capabilities, and new concepts are essential for successfully using the complete spectra of memristive functionalities for industrial applications. Here, we introduce two-terminal ohmic memristor, characterized by a different type of switching defined as filament conductivity change mechanism (FCM). The operation is based entirely on localized electrochemical redox reactions, resulting in essential advantages such as ultra-stable binary and analog switching, broad voltage stability window, high temperature stability, high switching ratio and good endurance. The multifunctional properties enabled by the FCM can be effectively used to overcome the catastrophic forgetting problem in conventional deep neural networks. Our findings represent an important milestone in resistive switching fundamentals and provide an effective approach for designing memristive system, expanding the horizon of functionalities and neuroscience applications.
format Article
id doaj-art-8e5234b352b140bda20d3ded15bd5052
institution OA Journals
issn 2041-1723
language English
publishDate 2025-03-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-8e5234b352b140bda20d3ded15bd50522025-08-20T01:57:52ZengNature PortfolioNature Communications2041-17232025-03-0116111310.1038/s41467-025-57543-wElectrochemical ohmic memristors for continual learningShaochuan Chen0Zhen Yang1Heinrich Hartmann2Astrid Besmehn3Yuchao Yang4Ilia Valov5Institute of Materials in Electrical Engineering 2 (IWE2), RWTH Aachen UniversityBeijing Advanced Innovation Center for Integrated Circuits, School of Integrated Circuits, Peking UniversityCentral Institute for Engineering, Electronics and Analytics (ZEA-3), Forschungszentrum JülichCentral Institute for Engineering, Electronics and Analytics (ZEA-3), Forschungszentrum JülichBeijing Advanced Innovation Center for Integrated Circuits, School of Integrated Circuits, Peking UniversityPeter Grünberg Institute 7 and JARA-FIT, Forschungszentrum JülichAbstract Developing versatile and reliable memristive devices is crucial for advancing future memory and computing architectures. The years of intensive research have still not reached and demonstrated their full horizon of capabilities, and new concepts are essential for successfully using the complete spectra of memristive functionalities for industrial applications. Here, we introduce two-terminal ohmic memristor, characterized by a different type of switching defined as filament conductivity change mechanism (FCM). The operation is based entirely on localized electrochemical redox reactions, resulting in essential advantages such as ultra-stable binary and analog switching, broad voltage stability window, high temperature stability, high switching ratio and good endurance. The multifunctional properties enabled by the FCM can be effectively used to overcome the catastrophic forgetting problem in conventional deep neural networks. Our findings represent an important milestone in resistive switching fundamentals and provide an effective approach for designing memristive system, expanding the horizon of functionalities and neuroscience applications.https://doi.org/10.1038/s41467-025-57543-w
spellingShingle Shaochuan Chen
Zhen Yang
Heinrich Hartmann
Astrid Besmehn
Yuchao Yang
Ilia Valov
Electrochemical ohmic memristors for continual learning
Nature Communications
title Electrochemical ohmic memristors for continual learning
title_full Electrochemical ohmic memristors for continual learning
title_fullStr Electrochemical ohmic memristors for continual learning
title_full_unstemmed Electrochemical ohmic memristors for continual learning
title_short Electrochemical ohmic memristors for continual learning
title_sort electrochemical ohmic memristors for continual learning
url https://doi.org/10.1038/s41467-025-57543-w
work_keys_str_mv AT shaochuanchen electrochemicalohmicmemristorsforcontinuallearning
AT zhenyang electrochemicalohmicmemristorsforcontinuallearning
AT heinrichhartmann electrochemicalohmicmemristorsforcontinuallearning
AT astridbesmehn electrochemicalohmicmemristorsforcontinuallearning
AT yuchaoyang electrochemicalohmicmemristorsforcontinuallearning
AT iliavalov electrochemicalohmicmemristorsforcontinuallearning